BUG #16168: Full text search parser prefers "file" over "asciiword"

The following bug has been logged on the website:

Bug reference:      16168
Logged by:          Chris Waters
Email address:      [email protected]
PostgreSQL version: 10.5
Operating system:   Any
Description:        

When parsing an English string to form a term vector the parser will treat
any words separated by / as a file path. This is a problem when / is used in
normal English, e.g.

select to_tsvector('morning/afternoon');
      to_tsvector      
-----------------------
 'morning/afternoon':1

I think it would be better if the parser output both the "file" and
"asciiword" versions of this string and indexed both.